PLEASE SCROLL DOWN FOR ARTICLE - Boston Universitysites.bu.edu/cliveg/files/2013/11/Robinson01.pdf · 2013. 11. 29. · PLEASE SCROLL DOWN FOR ARTICLE This article was downloaded

PLEASE SCROLL DOWN FOR ARTICLE

This article was downloaded by: [Duke University]On: 12 November 2008Access details: Access Details: [subscription number 788670354]Publisher Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

International Journal of Remote SensingPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713722504

An empirical approach to retrieving monthly evapotranspiration over AmazoniaNegrón R. I. Juárez a; M. L. Goulden b; R. B. Myneni c; R. Fu a; S. Bernardes d; H. Gao aa School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA, USA b EarthSystem Science and Ecology and Evolutionary Biology, University of California, Irvine, CA, USA c Departmentof Geography, Boston University, Boston, MA, USA d Department of Geography, University of Georgia,Athens, GA, USA

Online Publication Date: 01 December 2008

To cite this Article Juárez, Negrón R. I., Goulden, M. L., Myneni, R. B., Fu, R., Bernardes, S. and Gao, H.(2008)'An empirical approachto retrieving monthly evapotranspiration over Amazonia',International Journal of Remote Sensing,29:24,7045 — 7063To link to this Article: DOI: 10.1080/01431160802226026URL: http://dx.doi.org/10.1080/01431160802226026

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

http://www.informaworld.com/smpp/title~content=t713722504http://dx.doi.org/10.1080/01431160802226026http://www.informaworld.com/terms-and-conditions-of-access.pdf

An empirical approach to retrieving monthly evapotranspiration overAmazonia

R. I. NEGRÓN JUÁREZ*{, M. L. GOULDEN{, R. B. MYNENI§, R. FU{,S. BERNARDES" and H. GAO{

{School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta,GA, USA

{Earth System Science and Ecology and Evolutionary Biology, University of California,Irvine, CA, USA

§Department of Geography, Boston University, Boston, MA, USA

"Department of Geography, University of Georgia, Athens, GA, USA

(Received 30 April 2007; in final form 18 April 2008 )

The extent of evapotranspiration (ET) over the Brazilian Amazon rainforestremains uncertain because in situ measurement sites do not cover the entiredomain, and the fetch of these sites is only of the order of 103m. In thisinvestigation we developed an empirical method to estimate ET over the BrazilianLegal Amazon (BLA). The work was based on an improved physicalunderstanding of what controls ET over the Amazonia rainforest resulting fromanalyses of recent in situ observations. Satellite data used in this study include theEnhanced Vegetation Index (EVI) from the Moderate Resolution ImagingSpectroradiometer (MODIS) and the surface radiation budget from theInternational Satellite Cloud Climatology Project (ISCCP). The empirical modelwas validated by measurements performed at four upland forest sites. Theobserved values and the calculated modelled values at these sites had the samemean and variance. On a seasonal scale, regional modelled ET peaks during theaustral spring (September to November), as reported in the literature. Inaddition, the empirical model allows us to estimate the regional seasonal andinterannual distributions of ET/precipitation rates.

1. Introduction

Evapotranspiration (ET) is a key component that links climate to the terrestrialecosystem. At specific sites over the Amazon forest, ET contributes to about 50% ofthe total precipitation, as calculated by water balance methods and eddy correlationmeasurements (Salati 1987, Shuttleworth 1988). The geographical variation of thisrate remains unknown. Results from the Anglo-Brazilian Amazonian ClimateObservation Study (ABRACOS; Gash et al. 1996) and the more recent Large-ScaleBiosphere-Atmosphere Experiment in Amazonia (LBA) (LBA 1996, Avissar et al.2002, Keller et al. 2004) have provided a better understanding of the controls offorest ET at seasonal and interannual time scales. These studies have shown not onlya higher ET in the dry season than in the wet season but also a higher ET over areaswith less rainfall during the dry season in eastern and central Amazonia(Shuttleworth 1988, Nepstad et al. 1994, Malhi et al. 2002, Sommer et al. 2002,

*Corresponding author. Email: [email protected]

International Journal of Remote SensingVol. 29, No. 24, 20 December 2008, 7045–7063

International Journal of Remote SensingISSN 0143-1161 print/ISSN 1366-5901 online # 2008 Taylor & Francis

http://www.tandf.co.uk/journalsDOI: 10.1080/01431160802226026

Downloaded By: [Duke University] At: 16:07 12 November 2008

Souza Filho et al. 2005, Negrón Juárez et al. 2007). The maintenance of such a highrate of ET by the rainforest during the dry season plays a central role in determiningwhen the subsequent wet season onset will occur (Fu and Li 2004, Li and Fu 2004).A higher ET, as a result of the forests responding to increased solar radiation, canalso mitigate the impact of moderately dry anomalies on the surface climatecondition (Negrón Juárez et al. 2007). Therefore, estimation of basin-wide ET atseasonal to interannual scales is essential for determining seasonal and interannualclimate variability in the Amazonian region.

Current techniques used to estimate ET (e.g. the eddy correlation technique) arelimited to point measurements that represent ET within an area with a radius only ofthe order of 102 to 103m, and as long as the surface characteristics are the same.However, heterogeneities in land surface and soil characteristics mean thatextrapolation of point data beyond this area would be inaccurate because of thedynamic and regional variability of ET. In more remote areas, such as the westernAmazon forests, ET estimates from numerical models (Potter et al. 2001) are limitedby subjective assumptions that are yet to be validated. Satellite data offer analternative for estimation of ET over large areas by complementing previouslyobserved measurements and numerical simulations of ET. A common approach is torelate ET to remotely sensed surface temperature (Jackson et al. 1977, Hatfield 1983,Jackson 1988, Moran et al. 1989, Kustas 1990), solar irradiance (Jackson et al.1983), vegetation indexes (Di Bella et al. 2000, Nishida et al. 2003a,b, Loukas et al.2005, Nagler et al. 2005, Su et al. 2005) or energy balance (Blad and Rosenberg1974, Bastiaanssen et al. 2005) by physical and statistical/semiempirical methods orthe Penman–Monteith equation (PM; Monteith 1973, 1981).

Although these methods aim to provide the best possible estimate of ET, theirapplication to Amazon forests remains unknown because of the lack or inadequacy ofmeteorological variables required. For instance, calculation of stomatal conductance,an input in the PM equation, is frequently based on the work of Jarvis (1976), but inthis case, the main weakness is the assumption that environmental constraintsoperate independently (Monteith 1995). In addition, values of stomatal conductanceat different time scales are not easily comparable, as discussed by the AmericanSociety of Civil Engineers (ASCE) Evapotranspiration in Irrigation and HydrologyCommittee (Itenfisu et al. 2000,Walter et al. 2000). Although temperature can be usedto estimate ET over some areas in the Amazon, the relationship between surfacetemperature and ET varies in sign across Amazonia. In central-eastern Amazonia, theET is in phase with temperature (da Rocha et al. 2004, Hutyra et al. 2005), while it isout of phase in eastern Amazonia (Sommer et al. 2002). ET is also out of phase withtemperature during the wet season over southern Amazonia (Vourlitis et al. 2002).These observational results clearly suggest that a simple relationship between surfacetemperature and ET does not exist at the basin scale.

Most of the flux tower observations over Amazonian forests during the past twodecades have suggested that ET is primarily controlled by surface radiation andvegetation condition, consistent with that originally reported by Shuttleworth (1988).More recently, Yang et al. (2006) have shown that the Enhanced Vegetation Index(EVI; Justice et al. 1998, Huete et al. 2002) is the most important driver for estimatingET at a continental scale. Compared to the Normalized Difference Vegetation Index(NDVI), the EVI can better capture canopy structural variation, seasonal vegetationvariation, land cover variation, and biophysical variation for high biomass vegetationsuch as that in Amazonia (Gao et al. 2000, Huete et al. 2002). The EVI also performs

7046 R. I. Negrón Juárez et al.


well under heavy aerosol and biomass burning conditions, which are frequent in theregion (Miura et al. 1998). The aim of this work was to develop an empirical approachfor the estimation of ET based on the vegetation condition inferred from the EVI andsurface net radiation from the International Satellite Cloud Climatology Project(ISCCP; Zhang et al. 1995, 2004). The empirical model was trained and validated usingdata collected at eight upland forest sites over the Brazilian Legal Amazon (BLA).

2. Methodology

2.1 Study area and model construction

The study area encompassed the nine states of the BLA (figure 1) covering about56106 km2 (Brazilian Institute of Geography and Statistics, www.ibge.gov.br). Latentheat flux (lE) data from eight upland sites (table 1) from the LBA experiment were usedfor model construction and validation. Data from the K83 and RJA sites were used fortraining the empirical model, and data from the K67 andK34 sites were used for modelvalidation. Other sites (CRS, DCK, SIN and BRG) were used for model comparison.

Observed data show that lE has a strong linear relationship with measurements ofnet radiation (Rn) but at different rates from site to site. For example, figure 2 showsthe relationship between daily lE and Rn at sites K34, K83 and RJA. Latent heatflux represented ,70% of Rn at K34 and K83, but only 40% at the RJA site,indicating that, for the Amazonia, most of Rn is used to sustain ET. It is worthnoting that vapour pressure deficit (VPD) and wind speed can also influence ET(Shuttleworth 1988, da Rocha et al. 2004, Negrón Juárez et al. 2007).

Our empirical model was based on the assumption that ET over the Amazonianrainforests is primarily a function of EVI and Rn:

Figure 1. Study area and sites used. The background is the climatological (1961–1990)annual precipitation (mm) over the Brazilian Legal Amazon. For description of sites used seetable 1. The states are: AC, Acre; AM, Amazonas; AP, Amapá; MA, Maranhão; MT, MatoGrosso; PA, Pará; RO, Rondônia; RR, Roraima; TO, Tocantins.

Evapotranspiration over Amazonia 7047


Table 1. Sites used in this study.

Site South latitude (u) West longitude (u) Dry season Method Source (period of study)

Bragantina*(Igarapé-Açu, Belém, Pará)

BRG 1.1850 47.5706 Sep–Dec EB Sommer et al. (2002)(Apr 97–Mar 98)

Cuieiras Reserve(Manaus, Amazonas)

CRS 2.5894 60.1153 Jun–Aug EC Malhi et al. (2002)(Sep 95–Aug 96)

Cuieiras Reserve(Manaus, Amazonas)

K34 2.6090 60.2093 Jun–Aug EC Araújo et al. (2002){(Jul 99–Sep 00)

Ducke Reserve(Manaus, Amazonas)

DCK 2.9500 59.9500 Jun–Aug PM Shuttleworth (1988)(Sep 83–Sep 85)

Tapajós National Forest(Santarém, Pará)

K67 2.8853 54.9205 Jul–Dec EC Hutyra et al. (2007){(Jan 02–Jan 06)

Tapajós National Forest(Santarém, Pará)

K83 3.0502 54.9280 Jul–Dec EC da Rocha et al. (2004)§(Jul 00–Dec 02)

Biological Reserve of Jarú(Jı́-Paraná, Rondônia)

RJA 10.0784 61.9337 Jun–Aug EC von Randow et al. (2004){Negrón Juárez et al.(2007){ (Jan 00–Dec 02)

Sinop" (Sinop, Mato Grosso) SIN 11.4125 55.3250 Jun–Sep EC Vourlitis et al. (2002)(Aug 99–Jul 00)

EB, energy balance; EC, eddy covariance; PM, Penman–Monteith.*Secondary vegetation: its height changed from h52.3 (Apr 97) to h53.5m (Mar 98).{Data available at http://beija-flor.ornl.gov/lba/.{Data available at ftp://ftp.as.harvard.edu/pub/tapajos/k67_flux/.§Data available at http://www.ess.uci.edu/%7Elba/."Transitional (ecotonal) tropical forest.

7048R.I.Negrón

Juárezet

al.


ET~f Rn, EVIð Þ ð1Þ

where EVI is the EVI obtained from Moderate Resolution ImagingSpectroradiometer (MODIS) data. The theoretical basis that justifies our model iscomposed of two factors. First, EVI has been shown to be highly correlated tocanopy level CO2 uptake (Huete et al. 2006), and this uptake is closely related tocanopy conductance, gc (Dickinson et al. 1991, Sellers et al. 1996). Based on physicalprinciples, ET and gc are also highly correlated (Monteith 1973, 1981). Therefore, incombination, EVI and ET should be closely related across sites. Second, Rn is relatedto ET (figure 2) through physiological (mainly photosynthesis) and physicalprocesses. Additionally, research has shown that greater Rn increases leaftemperature, which, in turn, drives up the VPD.

Different functions were tested independently at each site. The model that bestrepresented the observed ET at all sites (ET_site, measured in mmday

21) was

ET site~C1zC2|EC3VI | RnISCCP{C4ð Þ ð2Þ

where RnISCCP (measured in Wm22) is the net radiation from the ISCCP (see section

2.3 for details) and C1, C2, C3 and C4 are constants.The percentage error between observed and calculated ET was calculated as

Error~+1

n

Xn

i

Y ið Þ{Y ’ ið Þj jY ið Þ

|100zs

!

ð3Þ

where Y and Y9 are the observed and calculated ET values, respectively, n is thenumber of elements in the series, and s is the standard deviation of errors. Constantsat K83 and RJA were adjusted to obtain the maximum determination coefficient(R2) between observed ET (ET_obs) and modelled ET (ET_grl) at these sitessimultaneously. ET_grl was then used to calculate ET for the whole BLA.

2.2 MODIS EVI

The MODIS EVI products (http://edcdaac.usgs.gov/main.asp) are derived fromsurface reflectance from the MODIS/Terra sensor, corrected for molecular scattering,ozone absorption and aerosols (Miura et al. 2001, Vermote et al.

Figure 2. Relationship between net radiation, Rn (Wm22), and the latent heat flux, lE,

(Wm22) at sites K34, K83 and RJA.



2002). The EVI is designed to optimize signals from vegetation and is sensitive in highbiomass regions (Justice et al. 1998). The MODIS EVI (EVI) was calculated from thered, near-infrared and blue reflectances (rred, rNIR and rblue, respectively) as:

EVI~G rNIR{rredð Þ= rNIRzarred{brbluezLð Þ ð4Þ

The coefficients used in the algorithm are L51, a56, b57.5 and G52.5, with a and brepresenting the aerosol resistance. Equation (4) uses the blue band to correct theaerosol influence in the red band similar to the Atmosphere Resistant Vegetation Index(ARVI; Kaufman and Tanré 1992). The soil correction coefficient, L, is derived fromthe Soil-Adjusted Vegetation Index (SAVI; Huete 1988).

Sixteen-day MODIS EVI images with resolution of 1 km (http://modis-land.gsfc.nasa.gov/) from 2000 to 2006 were processed. Monthly data were obtained byusing the number of 16-day average images that overlap the calendar monthweighted by the number of actual days that overlap that month. The time series ofmonthly images were then smoothed on a pixel-per-pixel basis using a three-pointcentral-moving-average. For regional estimation, the 1 km resolution data wereaggregated to 0.25u. For model construction and validation, a three-pixel boxcentred in the tower location was used. Figure 3 shows the monthly EVI at the K67,K83, K34 and RJA sites at 1 km and 0.25u resolutions for the period 2000–2004. Itwas verified that EVI values after aggregation had the same mean (t-Student, notshown) and variance (F-test, not shown).

Figure 3. Comparison of monthly EVI time series at 1 km and 0.25u horizontal resolution atK67, K83, K34 and RJA sites for the period 2000–2004.



2.3 ISCCP Rn data

Monthly net surface radiation (RnISCCP) was calculated for the period from July1983 to December 2004, at a spatial resolution of 2.5u. The calculation was based onsurface balance of shortwave (0.2–5.0 mm) and longwave (5.0–200 mm) radiative fluxdata at full-sky conditions from the ISCCP. Full-sky fluxes were weighted fromfluxes calculated for 15 types of cloudy conditions and clear-sky fluxes by theirfractional coverage for each cell. Zhang et al. (1995, 2004) have presented a completedescription of this data set.

For model construction and validation, we used the RnISCCP encompassing thetower coordinates. For the regional estimation of ET, the RnISCCP data wereinterpolated to a spatial resolution of 0.25u using the Kriging interpolation method(Oliver and Webster 1990). Figure 4 shows that RnISCCP at 0.25u and at 2.5u agreefairly well; however, an offset was observed when compared to the observed Rn. TheF-test statistics (and its probability) between observed Rn and RnISCCP at 0.25u forthe K67, K34, K83 and RJA sites was 1.074 (0.839), 1.569 (0.392), 1.245 (0.559) and2.667 (0.005), respectively. Except for the RJA site, RnISCCP at 0.25u had the samevariance with respect to the observed Rn. A detailed analysis of observed Rn andRnISCCP at RJA showed that the F-statistics for the period from January 2000 to

Figure 4. Comparison between observed net radiation, Rn_obs (Wm22), and net radiation

calculated from ISCCP data, RnISCCP, at 2.5u and 0.25u horizontal resolution at K67, K83,K34 and RJA sites.



December 2001 is 1.9, and its probability is 0.133, indicating that the observed andsatellite series had the same variance.

3. Results

3.1 Model training and validation

Figure 5 shows the observed and calculated ET at K83, RJA, K67 and K34. Theempirical formula trained for best fit at each specific site (ET_site) generally provideda close match with the observed data (ET_obs) at seasonal and interannual scales.However, at the RJA site, ET_site overestimated the annual maximum ET. Thedetermination coefficients (R2 at 95% CI, F-test) between ET_site and ET_obs were0.64 at the K83 site, 0.62 at the K67 site, 0.8 at the K34 site, and 0.32 at the RJA site.The general empirical model (ET_grl) used to calculate the ET over the BLA wasobtained using coefficients C152.7, C250.05, C351.75 and C45140. The R

2 values(95% CI, F-test) between ET_grl and ET_obs were 0.61, 0.55, 0.8 and 0.31 at the K83,K67, K34 and RJA sites, respectively. The associated errors between ET_site (ET_grl)and ET_obs at these sites were ¡17% (¡19%), ¡11% (¡13%), ¡6% (¡9%), and¡12% (¡20%), respectively. Using both RnISCCP and EVI at 0.25u spatialresolution, the ET_grl errors with respect to ET_obs at K67, K83, K34 and RJAwere ¡18%, ¡19%, ¡16% and ¡16%, respectively, with an average error of¡17%. This shows that the 0.25u resolution did not significantly diminish thequality of the calculated ET.

Figure 6 compares the seasonal mean of ET_grl for the period 2000–2004 at 1 kmand 0.25u resolutions with respect to ET_obs for the sites and periods listed in table 1.At the K34, K83, K67 and RJA sites, the average difference between ET_obs and

Figure 5. Comparison between observed, ET_obs (solid black line), and calculated values ofevapotranspiration using a calibrated model for each site, ET_site (grey line and open circles)and a general model based on K83 and RJA sites, ET_grl (grey line and filled squares), for sitesK83, RJA, K67 and K34.



ET_grl at 1 km of horizontal resolution was 8% during the wet season and 10% duringthe dry season. The differences at 0.25u resolution were very similar to the 1 km data,with a difference during the wet season of 8% and a difference during the dry seasonof 9%. At the CRS, DCK, BRG and SIN sites, the average difference between ET_obsand ET_grl during the wet season was 26% and 18% at 1 km and 0.25u resolutions,respectively. During the dry season these differences were 18% at 1 km resolutionand 19% at 0.25u resolution. The best agreement, approximately 11%, occurs at theCRS site in central equatorial Amazonia, located near the K34 site. The worstagreement, about 30%, is associated with the BRG site, located near the mouth ofthe Amazon River, where the climate is strongly influenced by wind coming fromthe ocean. Thus, on seasonal and interannual scales the maximum discrepancybetween ET_grl and ET_obs is about 30%, a value that is comparable to theuncertainties of in situ measurements of ET (e.g. see references in table 1).

Several factors can contribute to the discrepancies shown in figure 6. For instance,the satellite inputs associated with surface solar radiation and the EVI may bedifferent from those at the flux towers, in part due to the lack of information onsubscale structures within the satellite footprints. The observed values also havesignificant uncertainties due to instrumental errors, as well as the different methodsinvolved in the calculation of ET. For instance, the ET values at the DCK site wereobtained from a combination of observed and calculated data using the Penman–Monteith equation, while the BRG site ET was calculated using both the Bowenratio energy balance and the Penman–Monteith equation. The reported ET values atthe SIN site were obtained using the eddy correlation system and the Priestley–Taylor model (Priestley and Taylor 1972).

Figure 6. Observed (black bar) and calculated values of evapotranspiration (ET) at 1 km(light grey bar) and 0.25u (dark grey bar) of horizontal resolution during (a) the wet seasonand (b) the dry season. The observed ET was obtained for periods described in table 1 and thecalculated ET values correspond to the period 2000–2004 (except 2002). Sites in italics indicatethat the periods of observed ET values are out of the estimated period.



3.2 Seasonal ET

Figure 7 shows the average seasonal ET and precipitation for the period 2000–2004over BLA. Precipitation data are from the Tropical Rainfall Measuring Mission(TRMM) 3B43 product (http://trmm.gsfc.nasa.gov), which uses TRMM datamerged with other satellite data and available rain gauge records. This datasethas proven to be consistent with rain gauge data (Negrón Juárez et al. 2007) and hasthe advantage of a high spatial resolution. From December to February (DJF), themaximum values of ET (,3mmday21, figure 7(a)) appeared in the north of theAmazonas State, whereas precipitation was higher in southern BLA (.8mmday21,figure 7(b)). From March to May (MAM), ET did not have extreme spatial changesover BLA, showing an average value of 2.6mmday21, with some areas inMaranhão State having values of 2.9mmday21. During this period, precipitationwas concentrated over the northern BLA, reaching values higher than 6mmday21

(lower values were observed over the north of Roraima State, RR in figure 1). Theperiod from June to August (JJA) showed a gradient of precipitation (ET) fromsouth to north, with values ranging from 3mmday21 (2mmday21) to 7mmday21

(2.8mmday21). Maximum ET values of about 3mmday21 were observed in the

north of Pará State. From September to November (SON), ET ranged from 3 to3.3mmday21, except over the Cerrado domain (see figure 11) in the southern BLA,where ET ranged from 2.2 to 2.5mmday

21. In the eastern Amazonia ET was higherthan 3.1mmday21 when precipitation was lower than 3mmday21, which isconsistent with previous in situ observations (Nepstad et al. 1994, da Rocha et al.2004, Oliveira et al. 2005). The increase of ET in eastern Amazonia from Septemberto November is consistent with higher values of EVI and Rn, as observed by Hueteet al. (2006). However, the adjacent deforested areas presented lower values of ET aswell as of the EVI during the dry season, as reported in other studies (Huete et al.2006, Myneni et al. 2007).

Figure 7. Three-month average values (DJF,MAM, JJA, SON) and annual average values of(a) calculated ET (mmday

21) and (b) precipitation (mmday21), over the Brazilian LegalAmazon. Values were calculated for the period 2000–2004. Precipitation corresponds to theTRMM 3B43 data sets.



4. Discussion

4.1 Sensitivity of EVI to climate variability

In this paper, EVI is used to represent changes in vegetation, including the responseof vegetation to climate variability. It is important to ask how adequate EVI is inrepresenting the geographic seasonal to interannual variation of vegetation. TheAmazon forest is dominated by its response to rainfall variation on these time scales.The capability of EVI to monitor the seasonal variation of the Amazon forest wasstudied by Xiao et al. (2006). To study the capability of EVI to monitor theinterannual variability of Amazon forest we performed an analysis over an area inthe eastern Amazon (2u S–5u S, 53uW–56uW) that experiences regular droughtsfrom El Niño events (Stokstad 2005). In this area, the dry season lasts 5 months(consecutive months with precipitation ,100mm) centred in September (Sombroek2001). Droughts are one of the most important climate variability events in theAmazon, having a strong influence over vegetation phenology. We diagnose theseevents using the Standardized Precipitation Index (SPI; McKee et al. 1993),calculated for the period 1986–2006, and based on precipitation data from theGlobal Precipitation Climatological Centre (GPCC). The SPI value is equal to zeroif precipitation does not deviate from its climatology. Droughts begin when SPIvalues first fall below zero and end with positive or zero values of SPI for at leastseveral months.

Figure 8(a) shows monthly SPI time series from 2000 to 2006. During this period,drought events with different intensities can be observed in 2002, 2003 and 2005.The 2002–2003 drought, but not the 2005 drought, can be related to an El Niñoevent (McPhaden 2004, Marengo et al. 2008). It can also be observed that, exceptfor some months during the dry season, the precipitation in 2002 was below its long-term average, with a deficit starting around July 2001. Coincident with this deficit,figure 8(b) shows that, during the 2002 dry season, the EVI had the lowest valuesamong the drought events compared. This result agrees with Hutyra et al. (2007),who reported that the gross primary productivity for the period 2002–2005 waslower during the 2002 dry season in the K67 site. As EVI is sensitive to canopy

Figure 8. Monthly spatial average values over the area 2uS–5uS and 53uW–56uW of (a) theStandardized Precipitation Index (SPI) from 2000 to 2006 and (b) the dry season EVI from2002 to 2005. The SPI was calculated from the GPCC data (1986–2006) using a time scale of 6months.



structure, low EVI values can be associated with a thinning of the leaf canopy orcanopy openness as a consequence of drought conditions, as reported for this regionby Nepstad et al. (2002). In 2003, above average precipitation started in July andcontinued throughout July 2004. The increase of precipitation during the 2003 dryseason produced a favourable effect on vegetation (an increase in EVI values) withrespect to the same period in 2002. In 2005, South America experienced below-normal precipitation anomalies, but in west-central and eastern Brazil the rainyseason was slightly below normal (Shein 2006). In this year, after 6 months of wetterthan normal precipitation conditions, a moderate drought event was observed inSeptember. The EVI shows a peak in this month due to the positive response ofvegetation to well-watered soil and an increase in solar radiation.

Although vegetation green-up has been reported during the dry season in theeastern Amazon (Huete et al. 2002), our results show that the intensity of this green-up depends strongly on climate variability. The EVI was capable of monitoringforest conditions related to this variability.

4.2 Uncertainties of observed and calculated ET

First, in situ measurements of ET have uncertainties of about 10–30%, as suggestedby the observed imbalance of 14% (K83) to 28% (RJA) (see table 1 for references),related to several factors (e.g. advection, night flux underestimations).

Second, cloud cover can reduce the quality of MODIS EVI data. Asner (2001)reported that the chance of acquiring scenes with 30% or less cloud cover over theAmazon is minimal from December to May and very limited from October toNovember. Annually, the probability of obtaining scenes with 30% or less cloudcover in the BLA is .90% southwards of 5u S. The MODIS science team establishedthe Vegetation Index Usefulness Index (UI; http://edcdaac.usgs.gov/modis/moyd13_qa_v4.asp) as a quality assessment on a pixel-by-pixel basis. We used theUI to evaluate the quality of the EVI images. The values of the UI for a pixel aredetermined by several factors, including aerosol quantity, atmospheric correctionconditions, cloud cover, shadow, and sun-target-viewing geometry. The UI has 16levels that vary from perfect quality (level 0) to low quality (level 14). No useful dataare labelled as level 15. In this work the level ‘Intermediate Quality’ (sixth in the UIscale) was used as the threshold between good and bad quality pixels. This thresholdis more restrictive than the ‘Average Quality’ (eighth in the UI scale) commonlyused.

The 16-day 1 km61 km usefulness quality data were used to calculate thepercentage of pixels with quality better than the ‘Intermediate Quality’ level. Theresults were then converted to 0.25u resolution and are presented in figure 9. CentralPará had good quality EVI in 30–40% of images. This percentage was higher, near50%, over western Maranhão, southern Amapá, some areas in Roraima, and thewestern portion of Amazonas State. The remaining areas presented EVI better thanthe threshold in at least 60% of the cases, with the best EVI quality being observed inthe southern and western edges of the BLA. Our model represented the observationsat central Pará (K83 and K67 sites) well, overcoming the effects of pixel quality.However, on the western edge of BLA, the ET values should be considered withcaution because observed ET data are not available for comparison.

A third factor contributing to errors in ET are uncertainties in ISCCP-FD fluxes.The largest sources of uncertainties for South America are caused by clouds, TIROSOperational Vertical Sounder (TOVS) atmospheric temperature, ISCCP surface



temperatures, and vertical water vapour profiles (Zhang et al. 2004). Zhang et al.(2004) reported that the overall uncertainties of these fluxes are at least 10–15Wm22 at the surface. Although the differences between the site observations andthe ISCCP-FD fluxes presented here are larger than that, the differences appears tobe constant throughout the BLA (figure 4), presumably because of differences inspatial and temporal scales between satellite and flux tower measurement footprints.

4.3 ET/precipitation rates over the Amazon

The rate of ET determined with respect to precipitation across the Amazon forest isshown in figure 10. The DCK, CRS, K83 and SIN sites reported an average ET/precipitation rate of 50% and the RJA site had a rate of 45%. Our results agree withthese measurements but also show a regional variability in this rate. Areas ofmaximum precipitation (see also figure 1) had the lowest rate (,40%). These areasalso show from zero to three consecutive months with precipitation ,100mm andlower daylight intensity (Sombroek 2001). High precipitation implies high cloudiness,and therefore less radiation available to promote ET. These areas also have themaximum plant-available soil water (PAW) and are less susceptible to droughts thanthose observed during an El Niño event (Nepstad et al. 2004). ET/precipitation rateswith values greater than 40% were observed in southern, eastern and middle areas ofthe basin. These areas are characterized by annual precipitation,2000mm (figure 1),4–7 consecutive months of precipitation,100mm (Sombroek 2001) and low values ofmaximum PAW (Nepstad et al. 2004), and their temperatures strongly correlated with

Figure 9. Areas in the Brazilian Legal Amazon with recurrent problems of MODIS EVIquality. The grey-scale indicates the frequency that pixels have better than ‘IntermediateQuality’ over the course of a year.

Figure 10. Percentage of evapotranspiration with respect to precipitation over the BrazilianLegal Amazon for the period 2000–2004.



the El Niño-Southern Oscillation (ENSO) index (Malhi andWright 2004). As a result,these areas are more susceptible to drought events and fires.

4.4 ET over deforested areas

Since the early 1970s, the Amazon basin has been heavily deforested in an area alongthe southern and eastern edges of the basin, as delineated in figure 11. By August2004 the accumulated deforestation was ,14% of the BLA (www.obt.inpe.br/prodes). After 2002, an increase in the deforestation rate was observed in response tothe increased international demand for soybeans (especially in the state of MatoGrosso) and beef (Fearnside 2005). To determine whether calculated ET is adequateover deforested areas we focused our analysis over a small area within thecoordinates 10u S to 11u S and 55uW to 56uW in the Mato Grosso State, whereintense deforestation has occurred in recent years. During the wet and dry seasonour model shows ET values of 2.8¡0.13 and 2.2¡0.1mmday

21, respectively. Thesevalues are very close to those of 3.2mmday21 in December and 2.12mmday21 forthe dry season reported by Priante et al. (2004) over a cattle pasture area located at9.86u S and 55.23uW in the Mato Grosso State. These results suggest that our modelmay be used as a first approximation of ET over deforested areas.

5. Conclusions

The empirical model developed to calculate ET, using satellite-retrieved netradiation and EVI as inputs, agrees with in situ observations within 17%. Thisempirical model was based on the suggestion from previous in situ observations thatthe distribution of ET over the Amazon is largely controlled by the surface netradiative flux and vegetation condition. Our analysis suggests that the EVI canreasonably capture the vegetation responses due to rainfall variability on a largescale. The rates of ET/precipitation were very similar to those reported in theliterature over the BLA; although MODIS EVI data are reduced in quality overPará State, the calculated ET agreed well with observed ET, with high values over thenorthern part of this state during the dry season (September to November). Ourmodel enabled us to determine a regional pattern of the ET. The model also shows areasonable change in ET over deforested areas compared to in situ observations.However, further studies are needed to more clearly determine whether this methodcan be used to estimate ET over areas where change in land cover has occurred.

Figure 11. Distribution of forest (1) and savanna (2) across the Brazilian Legal Amazon.Area inside the solid line roughly represents areas with strong deforestation (source:www.obt.inpe.br/prodes/).



AcknowledgementsWe acknowledge the LBA community for both their extensive field measurements,which provide data focused on the understanding of the Amazon forest, and formaking these data available through the LBA Beija-Flor search and retrieval system(http://beija-flor.ornl.gov/lba). We thank Susan Ryan and John Trostel for theirvaluable editorial assistance. We are also grateful to the constructive commentsmade by the three anonymous reviewers. This work was supported by the NASATerrestrial Ecology Program for the Earth System Science Research Using Data andProducts from Terra and ACRIM Satellites.

ReferencesARAÚJO, A.C., NOBRE, A.D., KRUIJT, B., ELBERS, J.A., DALLAROSA, R., STEFANI, P., VON

RANDOW, C., MANZI, A.O., CULF, A.D., GASH, J.H., VALENTINI, R. and KABAT, P.,2002, Comparative measurements of carbon dioxide fluxes from two nearby towers ina central Amazonian rainforest: the Manaus LBA site. Journal of GeophysicalResearch-Atmospheres, 107, 8090, doi:10.1029/2001JD000676.

ASNER, G.P., 2001, Cloud cover in Landsat observations of the Brazilian Amazon.International Journal of Remote Sensing, 22, pp. 3855–3862.

AVISSAR, R., DIAS, P.L.S., DIAS, M.A.F.S. and NOBRE, C., 2002, The Large-Scale Biosphere-Atmosphere Experiment in Amazonia (LBA): insights and future research needs.Journal of Geophysical Research-Atmospheres, 107, 8086, doi:10.1029/2002JD002704.

BASTIAANSSEN, W.G., NOORDMAN, E.J., PELGRUM, H., DAVIDS, G., THORESON, B.P. andALLEN, R.G., 2005, SEBAL model with remote sensed data to improve water-resources management under actual field conditions. Journal of Irrigation andDrainage Engineering, 131, pp. 85–93.

BLAND, B.L. and ROSENBERG, N.J., 1974, Lysimetric calibration of the Bowen ratio-energybalance method for evapotranspiration estimation in the Central Great Plains.Journal of Applied Meteorology, 13, pp. 227–236.

DA ROCHA H.R., GOULDEN, M.L., MILLER, S.D., MENTON, M.C., PINTO, L.D.V.O., DEFREITAS H.C. and FIGUEIRA, A.M.E.S., 2004, Seasonality of water and heat fluxesover a tropical forest in eastern Amazonia. Ecological Applications, 14, pp. S22–S32.

DI BELLA, C.M., REBELLA, C.M. and PARUELO, J.M., 2000, Evapotranspiration estimatesusing NOAA AVHRR imagery in the Pampa region of Argentina. InternationalJournal of Remote Sensing, 21, pp. 791–797.

DICKINSON, R.E., HENDERSON-SELLERS, A., ROSENZWEIG, C. and SELLERS, P.J., 1991,Evapotranspiration models with canopy resistance for use in climate models: a review.Agricultural and Forest Meteorology, 54, pp. 373–388.

FEARNSIDE, P.M., 2005, Deforestation in Brazilian Amazonia: history, rates and con-sequences. Conservation Biology, 19, pp. 680–688.

FU, R. and LI, W., 2004, The influence of the land surface on the transition from dry to wetseason in Amazonia. Theoretical and Applied Climatology, 78, pp. 97–110.

GAO, X., HUETE, A.R., NI, W. and MIURA, T., 2000, Optical–biophysical relationships ofvegetation spectra without background contamination. Remote Sensing ofEnvironment, 74, pp. 609–620.

GASH, J.H.C., NOBRE, C.A., ROBERTS, J.M. and VICTORIA, R.L., 1996, An overview ofABRACOS. In Amazonian Deforestation and Climate, J.H.C. Gash, C.A. Nobre, J.M.Roberts and R.L. Victoria (Eds), pp. 1–14 (Chichester: John Wiley & Sons).

HATFIELD, J.L., 1983, Evapotranspiration obtained from remote sensing methods. InAdvances in Irrigation, D.E. Hilled (Ed.), pp. 395–416 (New York: Academic Press).

HUETE, A., DIDAN, K., MIURA, T., RODRIGUEZ, E.P., GAO, X. and FERREIRA, L.G., 2002,Overview of the radiometric and biophysical performance of the MODIS vegetationindices. Remote Sensing of Environment, 83, pp. 195–213.



HUETE, A.R., 1988, A soil-adjusted vegetation index (SAVI). Remote Sensing of Environment,25, pp. 295–309.

HUETE, A.R., DIDAN, K., SHIMABUKURO, Y.E., RATANA, P., SALESKA, S.R., HUTYRA, L.R.,YANG, W., NEMANI, R. and MYNENI, R., 2006, Amazon rainforest green-up withsunlight in dry season. Geophysical Research Letters, 33, L06405, doi:10.1029/2005GL025583.

HUTYRA, L.R., MUNGER, J.W., SALESKA, S.R., GOTTLIEB, E.W., DAUBE, B.C., DUNN, A.L.,AMARAL, D.F., CAMARGO, P.B. and WOFSY, S.C., 2007, Seasonal controls on theexchange of carbon and water in an Amazonian rain forest. Journal of GeophysicalResearch-Biogeosciences, 112, GO3008, doi: 10.1029/2006JG000365.

HUTYRA, L.R., MUNGER, J.W., NOBRE, C.A., SALESKA, S.R., VIEIRA, S.A. and WOFSY, S.C.,2005, Climatic variability and vegetation vulnerability in Amazonia. GeophysicalResearch Letters, 32, L24712, doi:10.1029/2005GL024981.

ITENFISU, D., ELLIOTT, R.L., ALLEN, R.G. and WALTER, I.A., 2000, Comparison of referenceevapotranspiration calculations across a range of climates. In Proceedings 4thDecennial National Irrigation Symposium, R.G. Evans, B.I. Benham and T.P. Trooien(Eds), pp. 216–227 (St Joseph, MI: American Society of Agricultural Engineers).

JACKSON, R.D., REGINATO, R.J. and IDSO, S.B., 1977, Wheat canopy temperature: a practicaltool for evaluating water requirements. Water Resources Research, 13, pp. 651–656.

JACKSON, R.D., 1988, Surface temperature and the surface energy balance. In Flow andTransport in the Nature Environment: Advances and Application, W.L. Steffen and O.T.Dermard (Eds), pp. 133–153 (Berlin: Springer-Verlag).

JACKSON, R.D., HATFIELD, J.L., REGINATO, R.J., IDSO, S.B. and PINTER, P.J. JR., 1983,Estimation of daily evapotranspiration from one time-of-day-measurements.Agricultural Water Management, 7, pp. 351–362.

JARVIS, P.G., 1976, The interpretation of the variations in leaf water potential and stomatalcontrol conductance found in canopies in the field. Philosophical Transactions of theRoyal Society of London, Series B, 273, pp. 593–610.

JUSTICE, C.O., VERMOTE, E., TOWNSHEND, J.R.G., DEFRIES, R., ROY, D.P., HALL, D.,SALOMONSON, V., PRIVETTE, J., RIGGS, G., STRAHLER, A., LUCHT, W., MYNENI, R.,KNYAZIKHIN, Y., RUNNING, S., NEMANI, R., WAN, Z., HUETE, A., VAN LEEUWEN W.,WOLFE, R., GIGLIO, L., MULLER, J., LEWIS, P. and BARNSLEY, M., 1998, TheModerate Resolution Imaging Spectroradiometer (MODIS): land remote sensing forglobal change research. IEEE Transactions on Geoscience and Remote Sensing, 36, pp.1228–1248.

KAUFMAN, Y.J. and TANRÉ, D., 1992, Atmospherically resistant vegetation index (ARVI) forEOS-MODIS. IEEE Transactions on Geoscience and Remote Sensing, 30, pp. 261–270.

KELLER, M., ALENCAR, A., ASNER, G.P., BRASWELL, B., BUSTAMANTE, M., DAVIDSON, E.,FELDPAUSCH, T., FERNANDES, E., GOULDEN, M., KABAT, P., KRUIJT, B., LUIZAO, F.,MILLER, S., MARKEWITZ, D., NOBRE, A.D., NOBRE, C.A., FILHO, N.P., DAROCHA H.,DIAS, P.L.D.S., vON RANDOW, C. and VOURILITIS, G., 2004, Ecological research inthe Large-Scale Biosphere-Atmosphere Experiment in Amazonia: early results.Ecological Applications, 14, pp. S3–S16.

KUSTAS, W.P., 1990, Estimates of evapotranspiration with a one- and two-layer model of heattransfer over partial canopy cover. Journal of Applied Meteorology, 29, pp. 704–715.

LBA, 1996, The Large-Scale Biosphere-Atmosphere Experiment in Amazonia: ConciseExperimental Plan, The LBA Science Planning Group. Document available atCPTEC/INPE, Cachoeira Paulista, SP, Brazil.

LI, W. and FU, R., 2004, Transition of the large-scale atmospheric and land surfaceconditions from the dry to the wet season over Amazonia as diagnosed by theECMWF reanalysis. Journal of Climate, 17, pp. 2637–2651.

LOUKAS, A., VASILIADES, L., DOMENIKIOTIS, C. and DALEZIOS, N.R., 2005, Basin-wide actualevapotranspiration estimation using NOAA/AVHRR satellite data. Physics andChemistry of the Earth, 30, pp. 69–79.



MALHI, Y., PEGORARO, E., NOBRE, A.D., PEREIRA, M.G.P., GRACE, J., CULF, A.D. andCLEMENT, R., 2002, Energy and water dynamics of a central Amazonian rain forest.Journal of Geophysical Research-Atmospheres, 107, 8061, doi:10.1029/2001JD000623.

MALHI, Y. and WRIGHT, J., 2004, Spatial patterns and recent trends in the climate of tropicalrainforest regions. Philosophical Transactions of the Royal Society of London. Series B,359, pp. 311–329.

MARENGO, J.A., NOBRE, C.A., TOMASELLA, J., OYAMA, M.D., DE OLIVEIRA G.S., DEOLIVEIRA R., CAMARGO, H., ALVES, L.M. and BROWN, I.F., 2008, The drought ofAmazonia in 2005. Journal of Climate, 21, pp. 495–516.

MCKEE, T.B., DOESKEN, N.J. and KLIEST, J., 1993, The relationship of drought frequencyand duration to time scales. In Eighth Conference on Applied Climatology, 17–22January 1993, Anaheim, CA, pp. 179–184 (Boston: Meteorological Society).

MCPHADEN, M., 2004, Evolution of the 2002/03 El Niño. Bulletin of the AmericanMeteorological Society, 85, pp. 677–695.

MIURA, T., HUETE, A.R., VAN LEEUWEN, W.J.D. and DIDAN, K., 1998, Vegetation detectionthrough smoke-filled AVIRIS images: an assessment using MODIS band passes.Journal of Geophysical Research, 103, pp. 32001–32011.

MIURA, T., HUETE, A.R., YOSHIOKA, H. and HOLBEN, B.N., 2001, An error and sensitivityanalysis of atmospheric resistant vegetation indices derived from dark target-basedatmospheric correction. Remote Sensing of Environment, 78, pp. 284–298.

MONTEITH, J.L., 1973, Principles of Environmental Physics (London: Edward Arnold).MONTEITH, J.L., 1981, Evaporation and surface temperature. Quarterly Journal of the Royal

Meteorological Society, 107, pp. 1–27.MONTEITH, J.L., 1995, Accommodation between transpiration vegetation and convective

boundary layer. Journal of Hydrology, 166, pp. 251–263.MORAN, M.S., JACKSON, R.D., RAYMOND, L.H., GAY, L.W. and SLATER, P.N., 1989,

Mapping surface energy balance components by combining Landsat ThematicMapper and ground-based meteorological data. Remote Sensing of Environment, 30,pp. 77–87.

MYNENI, R.B., YANG, W., NEMANI, R., HUETE, A., DICKINSON, R., KNYAZIKHIN, Y.,DIDAN, K., FU, R., NEGRÓN JUÁREZ, R., SAATCHI, S., HASHIMOTO, H., ICHII, K.,SHABANOV, N., TAN, B., RATANA, P., PRIVETTE, J., MORISETTE, J., VERMOTE, E.,ROY, D., WOLFE, R., FRIEDL, M., RUNNING, S., VOTAVA, P., EL-SALEOUS, N.,DEVADIGA, S., SU, Y. and SALOMONSON, V., 2007, Large seasonal swings in leaf areaof Amazon rainforests. Proceedings of the National Academy of Sciences, 104, pp.4820–4823.

NAGLER, P.L., CLEVERLY, J., GLENN, E., LAMPKIN, D., HUETE, A. and WAN, Z., 2005,Predicting riparian evapotranspiration from MODIS vegetation indices andmeteorological data. Remote Sensing of Environment, 94, pp. 17–30.

NEGRÓN JUÁREZ, R.I., HODNETT, M.G., FU, R., GOULDEN, M.L. and vON RANDOW, C.,2007, Control of dry season evapotranspiration over Amazonian forest as inferredfrom observations at a southern Amazon forest site. Journal of Climate, 20, pp.2827–2839.

NEGRÓN JUÁREZ, R.I., LI, W., FU, R., FERNANDES, K. and CARDOSO, A.O., 2008,Comparison of precipitation datasets over the tropical South American and Africancontinents. Journal of Hydrometeorology (in review).

NEPSTAD, D.C., DE CARVALHO C.R., DAVISON, E.A., JIPP, P.H., LEFEBVRE, P.A.,NEGREIROS, G.H., DA SILVA E.D., STONE, T.A., TREMBORE, S.E. and VIEIRA, S.,1994, The role of deep roots in the hydrological and carbon cycles of Amazonianforests and pastures. Nature, 372, pp. 666–669.

NEPSTAD, D.C., LEFEBVRE, P., DA SILVA, U.L., TOMASELLA, J., SCHLESINGER, P.,SOLORZANO, L., MOUTINHO, P., RAY, D. and BENITO, J.G., 2004, Amazon droughtand its implications for forest flammability and tree growth: a basin-wide analysis.Global Change Biology, 10, pp. 704–717.



NEPSTAD, D.C., MOUTINHO, P., DIAS, M.B., DAVIDSON, E., CARDINOT, G., MARKEWITZ, D.,FIGUEREDO, R., VIANNA, N., CHAMBERS, J., RAY, D., GUERREIROS, J.B.,LEFEBVRE, P., STERNBERG, L., MOREIRA, M., BARROS, L., ISHIDA, F.Y.,TOHLVER, I., BELK, E., KALIF, K. and SCHWALBE, K., 2002, The effects of partialthrough fall exclusion on canopy processes, aboveground production, and biogeo-chemistry of an Amazon forest. Journal of Geophysics Research, 107, 8085,doi:10.1029/2001JD000360.

NISHIDA, K., NEMANI, R.R., GLASSY, J.M. and RUNNING, S.W., 2003b, Development of anevapotranspiration index from Aqua/Modis for monitoring surface moisture status.IEEE Transactions on Geoscience and Remote Sensing, 41, pp. 493–501.

NISHIDA, K., NEMANI, R.R., RUNNING, S.W. and GLASSY, J.M., 2003a, An operationalremote sensing algorithm of land surface evaporation. Journal of GeophysicalResearch, 108, 4270, doi:10.1029/2002JD002062.

OLIVEIRA, R.S., DAWSON, T.E., BURGESS, S.S.O. and NEPSTAD, D.C., 2005, Hydraulicredistribution in three Amazonian trees. Oecologia, 145, pp. 354–363.

OLIVER, M.A. and WEBSTER, R., 1990, Kriging: a method of interpolation for geographicalinformation system. International Journal of Geographical Information Systems, 4, pp.313–332.

POTTER, C., DAVIDSON, E., NEPSTAD, D. and DE CARVALHO C.R., 2001, Ecosystem modelingand dynamic effects of deforestation on trace gas fluxes in Amazon tropical forest.Forest Ecology and Management, 152, pp. 97–117.

PRIANTE, N., VOURLITIS, G., HAYASHI, M., NOGUEIRA, J., CAMPELO, J., NUNES, P.,SOUZA, L., COUTO, E., HOEGER, W., RAITER, F., TRIENWEILER, J., MIRANDA, E.,PRIANTE, P., FRITZEN, C., LACERDA, M., PEREIRA, L., BIUDES, M., SULI, G.,SHIRAIWA, S., DO PAULO, S. and SILVEIRA, M., 2004, Comparison of the mass andenergy exchange of a pasture and a mature transitional tropical forest of the southernAmazon Basin during a seasonal transition. Global Change Biology, 10, pp. 863–876.

PRIESTLEY, C.H.B. and TAYLOR, R.J., 1972, On the assessment of surface heat flux andevaporation using large-scale parameters. Monthly Weather Review, 100, pp. 81–92.

SALATI, E., 1987, The forest and the hydrological cycle. In The Geophysiology of Amazonia:Vegetation and Climate Interactions, R.E. Dickinson (Ed.), pp. 273–296 (New York:John Wiley & Sons).

SELLERS, P.J., RANDALL, D.A., COLLATS, G.J., BERRY, J.A., FIELD, C.B., DAZLICH, D.A.,ZHANG, C., COLLELO, G.D. and BOUNOUA, L., 1996, A revised land surfaceparameterization (SiB2) for atmospheric GCMs, Part I: Model formulation. Journalof Climate, 9, pp. 676–705.

SHEIN, K.A. (Ed.), 2006, State of the climate in 2005. Bulletin of the American MeteorologicalSociety, 87, pp. S1–S102.

SHUTTLEWORTH, W.J., 1988, Evaporation from Amazonian rainforest. Proceedings of theRoyal Society of London, Series B. Biological Sciences, 233, pp. 321–346.

SOMBROEK, W., 2001, Spatial and temporal patterns of Amazon rainfall. Ambio, 30, pp.388–396.

SOMMER, R., SA, T.D.D., VIELHAUER, K., DE ARAUJO A.C., FOLSTER, H. and VLEK, P.L.G.,2002, Transpiration and canopy conductance of secondary vegetation in the easternAmazon. Agricultural and Forest Meteorology, 112, pp. 103–121.

SOUZA FILHO, J.D., RIBEIRO, A., COSTA, M.H. and COHEN, J.C., 2005, Control mechanismsof the seasonal variation of transpiration in a northeast Amazonia tropical rainforest[in Portuguese]. Acta Amazonica, 35, pp. 223–229.

STOKSTAD, E., 2005, Experimental drought predict grim future for rainforest. Science, 308,pp. 346–347.

SU, H.B., MCCABE, M.F., WORD, E.F., SU, Z. and PRUEGER, J.H., 2005, Modelingevapotranspiration during SMACEX: comparing two approaches for local andregional scale prediction. Journal of Hydrometeorology, 6, pp. 910–922.



VERMOTE, E.F., EL SALEOUS, N. and JUSTICE, C., 2002, Atmospheric correction of theMODIS data in the visible to middle infrared: first results. Remote Sensing ofEnvironment, 83, pp. 97–111.

vON RANDOW, C., MANZI, B., KRUIJT, A.O., DE OLIVEIRA P.J., ZANCHI, F.B., SILVA, R.L.,HODNETT, M.G., GASH, J.H.C., ELBERS, J.A., WATERLOO, M.J., CARDOSO, F.L. andKABAT, P., 2004, Comparative measurements and seasonal variations in energy andcarbon exchange over forest and pasture in South West Amazonia. Theoretical andApplied Climatology, 78, pp. 5–26.

VOURLITIS, G.L., PRIANTE, N., HAYASHI, M.M.S., NOGUEIRA, J.D., CASEIRO, F.T. andCAMPELO, J.H., 2002, Seasonal variations in the evapotranspiration of a transitionaltropical forest of Mato Grosso, Brazil. Water Resources Research, 38, 1094,doi:10.1029/2000WR000122.

WALTER, I.A., ALLEN, R.G., ELLIOTT, R., JENSEN, M.E., ITENFISU, D., MECHAM, B.,HOWELL, T.A., SNYDER, R., BROWN, P., ECHINGS, S., SPOFFORD, T.,HATTENDORF, M., CUENCA, R.H., WRIGHT, J.L. and MARTIN, D., 2000, ASCE’sstandardized reference evapotranspiration equation. In Proceedings 4th DecennialNational Irrigation Symposium, R.G. Evans, B.I. Benham and T.P. Trooien (Eds), pp.209–215 (St Joseph, MI: American Society of Agricultural Engineers).

XIAO, X., HAGEN, S., ZHANG, Q., KELLER, M. and MOORE, B., III, 2006, Detecting leafphenology of seasonally moist tropical forest in South America with multi-temporalMODIS images. Remote Sensing of Environment, 103, pp. 465–473.

YANG, F., WHITE, M.A., MICHAELIS, A.R., ICHII, K., HASHIMOTO, H., VOTAVA, P., ZHU,A.-X. and NEMANI, R.R., 2006, Prediction of continental-scale evapotranspiration bycombining MODIS and AmeriFlux data though support vector machine. IEEETransactions on Geosciences and Remote Sensing, 44, pp. 3452–3461.

ZHANG, Y., ROSSOW, W.B. and LACIS, A.A., 1995, Calculation of surface and top ofatmosphere radiative fluxes from physical quantities based on ISCCP data sets. Part1. Method and sensitivity to input data uncertainties. Journal of Geophysical Research,100, pp. 1149–1165.

ZHANG, Y., ROSSOW, W.B., LACIS, A.A., OINAS, V. and MISHCHENKO, M.I., 2004,Calculation of radiative fluxes from the surface to top of atmosphere based onISCCP and other global data sets: refinements of the radiative transfer model and theinput data. Journal of Geophysical Research, 109, D19105, doi:10.1029/2003JD004457.



Documents

PLEASE SCROLL DOWN FOR ARTICLE - Boston Universitysites.bu.edu/cliveg/files/2013/11/Robinson01.pdf · 2013. 11. 29. · PLEASE SCROLL DOWN FOR ARTICLE This article was downloaded